Development and testing of algorithms for home monitoring of freezing of gait in Parkinson’s Disease using wearable sensors

Freezing of gait (FOG) is a debilitating symptom affecting people with Parkinson's disease, characterised by a temporary inability to initiate or maintain walking. This study proposes a threshold-based algorithm using wearable sensors to detect freezing events in people with Parkinson's disease. The research includes laboratory experiments and long-term home monitoring to assess the performance of the algorithm in different settings. Individual thresholds are set for each subject based on the freeze index as key indicator of freezing episodes. Data are collected from waist-mounted and foot-mounted sensors in laboratory and home settings. The laboratory-based cohort consists of 30 subjects, including healthy individuals and those in intermediate and advanced stages of Parkinson's disease. The home monitoring cohort includes 16 subjects, including healthy individuals and those with mid-stage Parkinson's disease. The algorithm is validated by comparing the results with neurologists' reports based on video recordings. Reliable thresholds are determined for both waist and foot sensors, taking into account disease progression. The algorithm shows superior performance in the laboratory setting, with high sensitivity, specificity and accuracy. However, in the home monitoring setting, challenges associated with daily activities lead to a slight decrease in performance, while still achieving satisfactory results. These findings highlight the potential of wearable sensors for accurate detection of freezing events and underline the importance of personalised threshold settings. The study contributes to the development of Parkinson's disease monitoring systems by providing an objective assessment of freezing events. Future research could focus on develop algorithms to classify daily activities before the freezing detection, as well as establishing thresholds for advanced disease stages. Moreover, these findings have implications for improving quality of life and provide important considerations for optimising monitoring and treatment.